University of Groningen

Life-history traits of the Whiting polyploid line of the parasitoid Nasonia vitripennisLeung, Kelley; van de Zande, Louis; Beukeboom, Leo W.

Published in:Entomologia Experimentalis et Applicata

DOI:10.1111/eea.12808

IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite fromit. Please check the document version below.

Document VersionPublisher's PDF, also known as Version of record

Publication date:2019

Link to publication in University of Groningen/UMCG research database

Citation for published version (APA):Leung, K., van de Zande, L., & Beukeboom, L. W. (2019). Life-history traits of the Whiting polyploid line ofthe parasitoid Nasonia vitripennis. Entomologia Experimentalis et Applicata, 167(7), 655-669.https://doi.org/10.1111/eea.12808

CopyrightOther than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of theauthor(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).

Take-down policyIf you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediatelyand investigate your claim.

Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons thenumber of authors shown on this cover page is limited to 10 maximum.

Download date: 13-07-2020

SPEC IAL ISSUE : NEXT GENERAT ION B IOLOG ICAL CONTROL

Life-history traits of the Whiting polyploid line of theparasitoid Nasonia vitripennisKelley Leung* , Louis van de Zande & Leo W. BeukeboomGroningen Institute for Evolutionary Life Sciences, University of Groningen, PO Box 11103, 9700 CCGroningen, The

Netherlands

Accepted: 11 February 2019

Key words: diploidmale, triploid female, fitness, body size, parasitization rate, mate competition,

biocontrol, Hymenoptera, Pteromalidae

Abstract In hymenopterans, males are normally haploid (1n) and females diploid (2n), but individuals with

divergent ploidy levels are frequently found. In species with ‘complementary sex determination’

(CSD), increasing numbers of diploid males that are often infertile or unviable arise from inbreeding,

presenting a major impediment to biocontrol breeding. Non-CSD species, which are common in

some parasitoid wasp taxa, do not produce polyploids through inbreeding. Nevertheless, polyploidy

also occurs in non-CSDHymenoptera. As a first survey on the impacts of inbreeding and polyploidy

of non-CSD species, we investigate life-history traits of a long-term laboratory line of the parasitoid

Nasonia vitripennis (Walker) (Hymenoptera: Pteromalidae) (‘Whiting polyploid line’) in which

polyploids of both sexes (diploid males, triploid females) are viable and fertile. Diploid males pro-

duce diploid sperm and virgin triploid females produce haploid and diploid eggs. We found that

diploid males did not differ from haploid males with respect to body size, progeny size, mate compe-

tition, or lifespan. When diploid males were mated to many females (without accounting for mating

order), the females produced a relatively high proportion of male offspring, possibly indicating that

these males produce less sperm and/or have reduced sperm functionality. In triploid females, para-

sitization rate and fecundity were reduced and body size was slightly increased, but there was no effect

on lifespan. After one generation of outbreeding, lifespan as well as parasitization rate were increased,

and a body size difference was no longer apparent. This suggests that outbreeding has an effect on

traits observed in an inbred polyploidy background. Overall, these results indicate some phenotypic

detriments of non-CSD polyploids that must be taken into account in breeding.

Introduction

Polyploidy is the heritable condition of having more than

the typical number of chromosome sets. It is rare in the

animal kingdom but relatively frequent in insects. Poly-

ploid individuals have, for example, been reported from

the insect orders Coleoptera, Diptera, Embioptera, Hemi-

ptera, Hymenoptera, Lepidoptera, and Orthoptera (Otto

& Whitton, 2000). In particular, Hymenoptera are prone

to polyploidy, often as a consequence of a type of sex

determination that is common in the order, the

‘complimentary sex determination’ (CSD) mechanism

(Cook, 1993; van Wilgenburg et al., 2006; Heimpel & de

Boer, 2008). All hymenopterans have haplodiploid sex

determination. Males develop from unfertilized eggs and

have genomes of maternal origin only, whereas females

develop from fertilized eggs and inherit complete chromo-

some sets from both parents. Accordingly, males are typi-

cally haploid (1n) and females are diploid (2n). However,

under CSD, increasing numbers of diploid males can arise

from inbreeding as a result of homozygosity at csd loci.

Suchmales are often sterile, unviable, or sire sterile triploid

daughters (Stouthamer et al., 1992; van Wilgenburg et al.,

2006) which poses a burden on population growth (Zayed

& Packer, 2005) and hampers rearing programs for biolog-

ical control (Stouthamer et al., 1992).

Polyploids, in particular diploid males, have been

recorded for more than 80 species across the hymenop-

teran tree (van Wilgenburg et al., 2006; Heimpel & de

Boer, 2008; Harpur et al., 2013). In many of these cases,

the occurrence of polyploidy is linked to the CSD*Correspondence: E-mail: k.leung@rug.nl

© 2019 The Authors. Entomologia Experimentalis et Applicata published by JohnWiley & Sons Ltd

on behalf of Netherlands Entomological Society Entomologia Experimentalis et Applicata 167: 655–669, 2019 655This is an open access article under the terms of the Creative Commons Attribution License,which permits use, distribution and reproduction in anymedium, provided the original work is properly cited.

DOI: 10.1111/eea.12808

mechanism of sex determination. CSD occurs throughout

Hymenoptera, but is clearly absent in some major groups

(Stouthamer et al., 1992; van Wilgenburg et al., 2006;

Heimpel & de Boer, 2008; Asplen et al., 2009; Elias et al.,

2009). Individuals with CSD have either a single csd locus

ormultiple csd loci. Those that are heterozygous for at least

one csd locus develop into females, and those that are

homozygous or hemizygous for all csd loci develop into

males. Inbreeding causes loss of csd allelic diversity,

increases sterile homozygous diploid male production,

reduces population size, which in turn further increases

homozygosity. This results in a so called ‘diploid male vor-

tex’ and eventual extinction (Zayed & Packer, 2005; Hein

et al., 2009; Zaviezo et al., 2018). Exacerbating this prob-

lem is the inability of mating females to discriminate

against sterile diploid males (Harpur et al., 2013). Thus

CSD species may be more difficult to breed for biological

control because of this specific genetic basis of sex determi-

nation (Stouthamer et al., 1992).

Less is known about how polyploidy impacts non-

CSD species. Inbreeding does not cause polyploidy in

non-CSD species, and polyploidy in non-CSD species

may not lead to sterile diploid males (e.g., Whiting,

1960; Ma et al., 2015). The absence of CSD is not well

documented among the Hymenoptera, but it is espe-

cially prominent in the Chalcidoidea, Cynipoidea, and

Bethlyoidea parasitoid wasps (Cook, 1993; Cook & Cro-

zier, 1995; van Wilgenburg et al., 2006). As parasitoid

wasps are among the most widely used biocontrol agents

(van Lenteren et al., 1997; van Lenteren, 2012), knowl-

edge of the potential effects of inbreeding and polyploidy

in these taxa is important. Interestingly, polyploidy

could potentially also be advantageous for biological

control, that is, if polyploidy would confer some fitness

advantages, as for example larger yield or hardiness of

polyploids in plant breeding (Comai, 2005) and aquacul-

ture (Piferrer et al., 2009). To the authors’ knowledge

polyploidy has never been explored for beneficial effects

on insect breeding.

A case of non-sterile polyploidy exists in the parasitoid

wasp Nasonia vitripennis (Walker) (Hymenoptera: Ptero-

malidae), the most widely studied non-CSD parasitoid

wasp (Beukeboom & Desplan, 2003; Shuker et al., 2003;

Beukeboom & Kamping, 2006; Verhulst, 2010; Werren

et al., 2010; Verhulst et al., 2013). In rare instances poly-

ploidy has appeared spontaneously in N. vitripennis labo-

ratory stocks, although it has not been observed in wild

populations or collections (Whiting, 1960; Beukeboom &

Kamping, 2006). A ‘Whiting polyploid line’ (WPL) has

been maintained in the laboratory since the 1940s, and

used primarily for sex determination research (Whiting,

1960; Dobson & Tanouye, 1998; Beukeboom & Kamping,

2006; Beukeboom & van de Zande, 2010; Verhulst, 2010;

Verhulst et al., 2013). Diploid males are fertile, produce

diploid sperm and sire triploid females. Triploid females

have lowered fecundity owing to a high frequency of aneu-

ploid eggs, but they also produce enough viable euploid

(haploid and diploid) eggs to continue breeding (Whiting,

1960; Beukeboom&Kamping, 2006).

Despite its long history of research use, much of the

Whiting polyploidy strain’s baseline biology is unknown.

It is the first known available resource for studying

viable non-CSD polyploidy, and highly useful for inves-

tigating the potential practical advantages and disadvan-

tages of polyploidy in non-CSD hymenopterans. Both

non-polyploid and polyploid individuals can be reliably

generated for both sexes, and all combinations are viable

and capable of reproduction. Furthermore, the breeding

scheme of the WPL suggests how non-CSD polyploids

may be used in a biocontrol breeding program. That is,

if the female polyploid state confers something beneficial

to biological control, this can be reliably passed on

through diploid males and a large number of polyploid

females can be produced as active biological agents every

other generation. Like most commercial hymenopteran

lines, the WPL has also become highly inbred over time,

so differences between non-polyploids and polyploids

within this strain can be attributed to ploidy state alone,

without genetic variation being a contributing factor.

These attributes make the WPL a suitable model for a

first survey of how polyploidy effects non-CSD para-

sitoid wasp breeding.

In this study, we compare life-history traits of poly-

ploids and non-polyploids of both sexes from the N. vit-

ripennis WPL. Specifically, we compare female and male

lifespan under starvation and feeding conditions, male

mate competition ability, progeny size and sex ratio, and

female parasitization ability. We do this for individuals of

both sexes from the long-used inbred maintenance

scheme, as well as for females after a single generation of

outbreeding. The latter is a first attempt to discriminate

between inbreeding vs. polyploidy effects. In previous

studies on other parasitoid wasp species, polyploids were

disadvantaged for some traits such as sterility (Stouthamer

et al., 1992; van Wilgenburg et al., 2006; Elias et al., 2009)

but not for others such as lifespan (Clark & Rubin, 1961)

and male mate competition (Elias et al., 2009; Harpur

et al., 2013; Thiel et al., 2014), although these studies were

exclusively on CSD species. We therefore anticipated that

the WPL polyploids might underperform non-polyploids

in some traits, although we could not predict which. We

discuss the results in the context of the significance of

polyploids in breeding and their performance as biocon-

trol agents.

656 Leung et al.

Materials and methods

Nasonia culture and crosses

Nasonia wasps were cultured at 25 °C, ca. 55% r.h., and

L16:D8 light cycle and hosted on commercially produced,

purchased Calliphora sp. pupae (Titus Blom, Groningen,

The Netherlands). The Whiting polyploid line (WPL)

originated spontaneously in Whiting’s cultures and has

been maintained in the laboratory since the 1950s. It was

acquired from the John H. Werren laboratory (University

of Rochester, Rochester, NY, USA) and has been kept in

diapause in our laboratory for 10 years with about one

generation of breeding every year. In active culture, the

strain is maintained following previously described proto-

cols (Whiting, 1960; Dobson & Tanouye, 1998; Beuke-

boom & Kamping, 2006). It carries complementary eye

color markers oyster (oy) and scarlet (st). Homozygotes or

hemizygotes for oy and st have gray and red eyes, respec-

tively, whereas wildtype eyes are dark purple. Wildtype-

eyed triploid females (oy +/ + st/ + st) are hosted (are given

hosts) as virgins and produce four types of sons: gray-eyed

haploids (oy +), red-eyed haploids (+ st), red-eyed diploids

(+ st/ + st), and wildtype-eyed diploids (oy +/ + st) (Fig-

ure 1). About twice as many haploids as diploids are pro-

duced, possibly due to embryonic lethality (Whiting,

1960). To continue the line, diploid wildtype-eyed males

(oy+/ + st) are mated to virgin (+ st/ + st) females from the

same inbred red-eye marker stock line used to originate

the WPL, scarlet. The resultant triploid daughters (oy +/ +st/ + st) are used to re-start the breeding cycle (Beukeboom

&Kamping, 2006).

To eliminate possible effects of generation and different

eye color markers, we compared life-history traits of hap-

loid and diploid males in the same generation from the

crossing scheme described above. In the male assays for

lifespan, progeny size, and offspring sex ratio, only the

red-eyed males were used as these can be haploid or

diploid (Figure 1). However, in the mate competition

assays, red-eyed haploids and dark wildtype-eyed diploids

were used to easily distinguish between the ploidy levels. It

is possible that red-eyed males have impaired vision, but a

pilot study established that the red-eye phenotype does

not impact the ability of males to acquire mates. For assays

comparing females, diploid and triploid females from the

same generation were reared on the same host batch to

prevent confounding effects of parental age and host qual-

ity. To generate these females, red-eyed haploid (+ st) and

red-eyed diploid (+ st/ + st) WPLmales of the same gener-

ation were mated to virgin females of the mutant scarlet

line. In a separate experiment to test for effects of inbreed-

ing, these males were also mated to isogenic females of the

AsymCx line derived from The Netherlands and cured of

Wolbachia bacteria (Breeuwer & Werren, 1990; Werren

et al, 2010). The WPL/scarlet inbred cross (from here on

‘WPL-inbred’) produced diploid (+ st/ + st) and triploid

(+ st/ + st/ + st) daughters, both with red eyes. The WPL/

AsymCx outbred cross (from here on ‘WPL-outbred’)

resulted in diploid (+ st/ + +) and triploid (+ st/ + st/ + +)daughters, which both have wildtype eyes, and were

assessed for the same assays as the WPL. Diploid and tri-

ploid daughters within each cross were compared for lifes-

pan and parasitization rate. The individuals used in each

assay (polyploid and non-polyploid counterparts that

were compared against each other), their eyemarker geno-

types and phenotypes, and the crosses used to obtain them

are described in Figure 1.

Ploidy level typing

Ploidy level was determined either through the virgin

daughter offspring count method or the flow cytometry

method. The virgin daughter method was used for lifespan

assays because the flow cytometry method requires the

male to be freshly killed. For the virgin daughter offspring

count method, a WPL male was given 24 h to mate with a

virgin AsymCx female and the female was subsequently

provided with two hosts. A pilot study established that

there is no lifespan difference between males mated

through this method and unmated males. Approximately

2 weeks later, daughters were collected as black pupae and

allowed to eclose. A virgin daughter of each cross was given

two hosts and their offspring were allowed to develop to

adulthood and emerge from hosts naturally. Triploids

have low fecundity due to egg aneuploidy (given that

Nasonia have five chromosomes, approximately one out

of 32 eggs are euploid and viable). A pilot study found that

scarlet virgin diploid females produce on average

29.00 � 25.68 (mean � SD; n = 45) offspring with two

hosts, but virgin triploid WPL females produce only

2.60 � 2.35 offspring (n = 40) (with the maximum being

nine). Therefore, in this study, if fewer than eight offspring

from the two hosts emerged after 2 weeks, the female was

assigned triploid status and her experimental father was

assigned diploid status. If more than 15 offspring emerged,

the female was assigned diploid status and her father was

considered a haploid male. Those females that had 9–15offspring could not be unambiguously typed for ploidy.

This applied to <5% of females and they were excluded

from analyses.

The flow cytometry method was used to type the ploidy

of males in the progeny size and progeny sex ratio assays,

one trial of the mate competitions, and the daughters of

themate competitions. It was adapted from previous poly-

ploid wasp studies (Beukeboom et al., 2007; Thiel et al.,

2014). Heads of wasps were removed with a razorblade

Whiting polyploid line ofNasonia vitripennis 657

and stored at �20 °C in individual 1.5-ml Eppendorf

tubes. Only heads were used for flow cytometry as ganglial

cells have consistent ploidy but body cells can be

endopolyploid (Fankhauser, 1945; Wertheim et al., 2013).

Upon preparation for flow cytometry, each head was

ground for 30 s with an electric VWR pellet mixer in

500 ll ice-cold Galbraith buffer [21 mM MgCl2, 30 mM

tri-sodium citrate dehydrate, 20 mM 3-[N-morpholino]

propane sulfonic acid (=MOPS), 0.1% Triton X-100, and

1 mg l�1 RNase A]. Each sample was then poured

through a Falcon 5-ml tube with a cell strainer cap to

remove exoskeletal material from the cell solute. Twenty

lg of propidium iodide (10 ll of a 2.5 mg ml�1 solution)

(Sigma-Aldrich, St. Louis, MO, USA) was added to each

sample, and the tube gently flicked to mix. Samples were

put on ice and immediately run through a BDFACS ARIA

II flow cytometer. Individuals were assigned ploidy based

on how their signal matched those of reference 1n male

(AsymCx), 2n female (AsymCx), and 3n female (WPL

females) samples. Individuals that had an ambiguous

FACS reading (e.g., a large amount of debris) were

excluded from analyses (ca. 10% of samples).

+ + + +

Gray 1n WPL (not used)

+ st + st

oy + + st + st

Wildtype 3n WPL

+ + + +

+ st + st

+ st + st oy +

+ st

Wildtype 2n AsymCx

Red 2n scarlet

Wildtype 2n AsymCx

Red 2n scarlet

Red 2n scarlet

Wildtype 3n WPL (line maintenance)

Red 2n WPL/scarlet

(inbred)

+ st + st + st

Red 3n WPL/scarlet

(inbred)

+ st + st + +

+ st + st oy +

-Headwidth -Starvation/fed lifespan

-Single/multi-mate competition for HVRx virgin females

-Progeny size and sex ratio

-Head width -Starvation/fed lifespan

-Parasitisation rate

-Head width -Starvation/fed lifespan

-Parasitisation rate

Wildtype 2n WPL

Red 2n WPL Red

1n WPL

+ st

Wildtype 3n WPL/AsymCx

(outbred)

+ st + st

oy + + st + st

+ st + +

Wildtype 2n WPL/AsymCx

(outbred)

Figure 1 Crossing scheme and individuals used in life-history trait assays with descriptors of sex, eye color, ploidy level (in bold), and

background. Vertical lines indicate descent and horizontal lines indicate a cross. Individuals withWhiting polyploid line (WPL)

background have descriptors above their sexual symbols, and unrelated lines used for crossing have descriptors below their sexual symbols.

Solid gray arcs indicate non-polyploid and polyploid counterpart pairings that were compared in life-history assays. Gray text indicates

which life-history traits were measured in these pairings.

658 Leung et al.

Lifespan

Lifespan was measured for the WPL red-eyed males (hap-

loid and diploid) and the WPL-inbred and WPL-outbred

females (diploid and triploid). Following parasitization,

cultures were checked for emerged offspring every day

starting at day 10 after oviposition. Upon emergence of the

first wasp, host pupae were opened and all individuals

were collected. Only individuals fully eclosed from the

pupal case and free walking were used for lifespan assays.

Each individual was housed in a 63 9 11 mm tube and

returned to standard rearing conditions. Every 12 h, all

tubes were checked for dead wasps. A wasp was scored as

dead if it was found at the bottom of the tube and was

unresponsive to gentle prodding with a fine-tip brush.

These assays were conducted under both starvation and

feeding conditions. Under feeding conditions, a strip of fil-

ter paper dipped in 10% sucrose solution was placed into

the tube and replaced every 3 days.

Male mate competition

Male mate competition assays were conducted to compare

haploid vs. diploid male ability to acquire female mates,

for both single- and multiple-mate availability scenarios.

From a hosting of virgin female triploid WPL, males were

collected and sorted by eye color. In each mate competi-

tion setup, a red-eyed male (presumed haploid, as about

75% of red-eyed males are haploid; Whiting, 1960) and a

wildtype-eyed male (diploid), randomly chosen and less

than 1 day old, were placed in a 63 9 11 mm tube.

Whereas red-eyed haploid and red-eyed diploid competi-

tion pairs would have been ideal to exclude effect of eye

color, the probability of randomly chosen pairs being one

haploid and one diploid was too low. In contrast, the

diploidy of males with wildtype eyes is known a priori.

Red-eyed haploids instead of oyster-eyed haploids were

used because too few of the latter were produced. In a pilot

study using only haploid males, females did not discrimi-

nate against oyster-eyed (from the pure oyster line) or red-

eyed (from the pure scarlet line) males when given a choice

between an eye-color mutant or a wildtype-eyed AsymCx

male. Therefore, eye color is not likely to be a factor in

male mate competition success or biased results here.

In the male mate competition setups, 1-day-old virgin

female mates were used from the genetically variable labo-

ratory population HVRx (van de Zande et al., 2014). This

population was chosen to better test male ability to attract

mates of genetically variable backgrounds, and to avoid

female re-mating behavior. Female polyandry evolves in

inbred and long-established laboratory lines (van den

Assem & Jachmann, 1999; Burton-Chellew et al., 2007;

Shuker et al., 2007), but the HVRx population is recently

derived from wild populations, maintained to prevent

selection for laboratory traits, and has not been bred

beyond 150 generations. Thus, females of this population

are likely to be strongly monandrous like wild Nasonia

(van den Assem et al., 1980; Burton-Chellew et al., 2007;

Grillenberger et al., 2008). For single-mate competition,

one HVRx virgin female was provided to the competing

haploid male and diploid male (n = 22 trials), and in the

multiple-mate competition setups, 10 HVRx virgin

females were provided (n = 20 trials). All setups were

given 24 h to mate. Afterwards, all males were frozen and

stored at�20 °C.Female mates were hosted individually and after

2 weeks, all offspring were collected and the larval (imma-

ture), male, and female offspring were counted for each

mate. Total family size (including larvae) and offspring sex

ratio (disregarding the few larvae, as they cannot be sexed)

were recorded. Offspring were then stored at �20 °C for

the follow-up flow cytometry analysis, using a single

daughter of each female. If a daughter was typed as diploid,

it was assumed that this was the ploidy of all her sisters,

and that her father (the female’s mate) was the haploid

red-eye male. Conversely, if a daughter was typed triploid,

it was assumed that she and all her sisters were sired by the

diploid dark-eyed male. In the unlikely case that multiple-

sired progeny would have occurred, the chance to score it

as fathered by a haploid or diploid male would have been

equal, so this would not bias the results. Females that pro-

duced all-male offspring were scored as notmated and dis-

carded from the analysis. Multiple-mate trials in which a

majority of the females did not mate were also discarded.

If all mated females of a multiple-mate trial were found to

have mated with a diploid mate, this was interpreted to

mean that it was possible that the red-eyed male and the

wildtype-eyed male were both diploid as red-eyed males

can also be diploid. In the single case in which this

occurred, to confirm that a haploid male vs. a diploid male

competition took place, the frozen red-eyed male was pro-

cessed with flow cytometry and its haploid status was veri-

fied.

Progeny size and sex ratio

Although only female offspring reflect a male sire, signals

from male mates may potentially influence female deci-

sions about total family size and offspring sex ratio. There-

fore, the total number of offspring (progeny size,

including both males and females) and offspring sex ratio

(proportion males out of total offspring) data were col-

lected for the AsymCx female mates ofWPL red-eyed hap-

loid and red-eyed diploid males used in lifespan assays.

The same data were collected for HVRx females in the

mate competition trials that used haploid red-eyed males

and diploid wildtyped-eyedmales.

Whiting polyploid line ofNasonia vitripennis 659

Female parasitization rate

Diploid and triploid virgin females of theWPL-inbred and

WPL-outbred lines were tested for their relative ability to

parasitize an excess of hosts. One day post pupal eclosion,

each female was given a first set of 10 hosts in a

63 9 11 mm tube, kept under standard conditions. Every

2 days, they were each transferred to another 10 fresh

hosts, and the previous hosting sets were allowed to

develop. This was repeated for the entirety of the speci-

men’s lifetime. After 2 weeks the hosts were scored for fly

or wasp emergences. If a fly emerged, it was scored as a

failed parasitization. If neither a fly nor wasps emerged it

was scored as parasitized (i.e., stung by the wasp with or

without oviposition). It is possible that for some of these

hosts, non-emergence could be attributed to bad host

quality rather than parasitization, but the proportion of

bad-quality hosts in our culturing is typically very low

(<5%) and considered to have negligible effect on results.

Deaths were also recorded every time females were re-

hosted, tomeasure approximate lifespan.

Head width (body size)

Head width data were collected to infer whether body size

differs between non-polyploids and polyploids, and

whether it is a factor in any significant differences for life-

history trait phenotypes as is often the case with insects

(Beukeboom, 2018). Whereas gregarious parasitoids such

asNasonia spp. can vary greatly in size based on host effect

(e.g., genetically identical males can vary by a factor of 2;

Groothuis & Smid, 2017), the host batch was the same for

directly compared datasets and host effect on body size

was assumed to be insignificant due to random use. As the

abdomen size can fluctuate over a specimen’s lifetime

from feeding or egg load, head width was used as a proxy

for total body size (as in, e.g., Charnov & Skinner, 1984).

Head width was measured for a subset of haploid and

diploid males of the WPL, and a subset of diploid females

and triploid females of the Whiting inbred and Whiting

outbred lines. Measurements were taken of heads sepa-

rated from bodies at 59 magnification with a Carl Zeiss

Stemi 508 dissection microscope with a W-PI 109/23 eye-

piece and 14 mm reticule.

Statistical analysis

Statistical tests were performed in SPSS Statistics v.25

(IBM, 2017) and RStudio v.1.0.153 (R Core Team, 2014)

with significance level a = 0.05. Shapiro–Wilk tests and

Levene tests were used to test for normality and homo-

geneity of variance for all datasets. A two-sample t-test was

used if data were normally distributed and had equal vari-

ance. A Welch t-test was used if distributions were normal

but variances were unequal. Mann–Whitney U tests were

used if data were neither normal nor equal in variance. A

binomial test was used to test for significant difference in

the mate competition ability of haploids vs. diploid males,

with the null hypothesis (H0) that haploid males and

diploid males have equal mate competition ability (each

mated with half of the females for both the individual- and

multiple-mate trials). Additionally, a general linear mixed

model was used to test whether females were more likely

to mate with a haploid or a diploid male in the single-mate

and multiple-mate experiments. This used a multinomial

logistic regression with a generalized logit link, with trial

number set as a random effect. A Satterthwaite approxi-

mation and an estimation of robust variance were used to

account for non-normality and low sample size. Survival

graphs were generated for all lifespan datasets of corre-

sponding non-polyploids and polyploids (for starvation,

feeding, and female parasitization rate) and analyzed with

the log-rank (Mantel-Cox) test to compare survival distri-

butions. To test for diploid vs. triploid parasitization abil-

ity, a binomial general linearized mixed model was used

for the number of hosts parasitized per host set using a

binary logistic regression link. For this, day and individual

were set as random effects, and ploidy state (diploid or tri-

ploid) as a fixed effect. Again, a Satterthwaite approxima-

tion and an estimation of robust variance were used. For

brevity, most data means, sample sizes, and P-values of

tests are reported in Table 1. The general linearized models

and binomial general mixed models are reported in the

Supporting Information.

Results

Lifespan

Polyploids generally do not live longer or shorter than

non-polyploids. This is true for diploid males and triploid

females of the Whiting inbred line, but not the triploid

Whiting outbred line females (Figure 2). The average lifes-

pan of starved WPL-inbred red-eyed haploid males was

similar to that of the red-eyed diploids (Table 1), and their

survival distributions were similar (log-rank test:

v2 = 0.167, d.f. = 1, P = 0.68) (Figure 2A). When fed,

there was also no difference in lifespan between red-eyed

haploid and diploid males (Table 1) and survival distribu-

tions were similar (log-rank test: v2 = 0.050, d.f. = 1,

P = 0.82) (Figure 2C). As body size (i.e., head width) was

not significantly different between haploid and diploid

males (see below), body size differences did not factor in

these results.

Under starvation conditions, diploid WPL-inbred

females had the same lifespan as triploid WPL-inbred

females (Table 1), and survival distribution was also the

same (log-rank test: v2 = 0.856, d.f. = 1, P = 0.36)

660 Leung et al.

(Figure 2B).When fed, the average lifespan of diploids and

triploid females of this line again did not differ (Table 1)

nor were their survival distributions different (log-rank

test: v2 = 0.801, d.f. = 1, P = 0.65) (Figure 2D). Starved

diploid WPL-outbred females lived 0.32 days less than tri-

ploid WPL-outbred females (Figure 2B). This difference,

although small, is significant (Table 1), as is the difference

in survival distribution (log-rank test: v2 = 4.231,

d.f. = 1, P = 0.04) (Figure 2B). In the fed assay,WPL-out-

bred diploid females lived a significant 2.1 days longer

than WPL-outbred triploids (Table 1, Figure 2D) and sur-

vival distributions are significantly different (log-rank test:

v2 = 9.035, d.f. = 1, P = 0.003) (Figure 2D). Thus, there

were no significant lifespan differences in the original

WPL-inbred genetic background for either sex. However,

in the outbred WPL background polyploid females lived

slightly longer than the non-polyploids under starvation

conditions, but under fed conditions, they livedmore than

2 days shorter than the non-polyploids.

Male mate competition

There is no difference in mate competition ability between

haploid and diploid males for either single female or mul-

tiple female mates. In the single-mate competition assay,

in 18 out of 22 trials, the HVRx female mated with a male

and produced daughters (the other four trials resulted in

Table 1 Data summary of life-history assays. Unless otherwise noted, data are means (� SD) of traits for non-polyploids (male haploids,

female diploids) and polyploids (male diploids, female triploids) of the Whiting polyploid line (WPL)-inbred and WPL-outbred back-

grounds. Assays with significant differences (P<0.05) between the non-polyploid and polyploids are indicated with an asterisk

Trait Assay Non-polyploid (n) Polyploid (n) P Test

Lifespan

(starved) (days)

WPL-inbredmale 5.66 � 2.82 (126) 5.75 � 1.88 (20) 0.34 Mann–Whitney U

WPL-inbred female 5.61 � 1.05 (42) 5.25 � 1.13 (12) 0.48 Mann–Whitney U

WPL-outbred female* 3.46 � 0.89 (80) 3.78 � 0.79 (67) 0.037 Mann–Whitney U

Lifespan (fed

10% sucrose

solution) (days)

WPL-inbredmale 22.21 � 8.79 (71) 23.33 � 10.55 (20) 0.78 Mann–Whitney U

WPL-inbred female 12.37 � 1.13 (41) 12.59 � 9 (12) 0.51 Mann–Whitney U

WPL-outbred female* 14.05 � 9.09 (228) 12.39 � 7.22 (114) 0.009 Mann–Whitney U

Malemate

competition

(trials won)

WPL-inbredmale (single

competition)

11 7 0.48 Binomial

WPL-inbredmale (multi

competition) (3 ties)

6 6 0.78 Binomial

Progeny size

(total, male and

female)

WPL-inbredmale (no choice) 49.30 � 13.15 (56) 49.26 � 8.18 (23) 0.99 Welch’s t-test

WPL-inbredmale (single

competition)

54.18 � 10.2 (11) 49.42 � 8.43 (7) 0.35 Two sample t-test

WPL-inbredmale (multi

competition)

54.97 � 14.63 (71) 55.07 � 15.57 (59) 0.97 Two sample t-test

Progeny sex ratio

(male/total)

WPL-inbredmale (no choice 0.147 � 0.08 (56) 0.181 � 0.11 (23) 0.16 Mann–Whitney U

WPL-inbredmale (single

competition)

0.33 � 0.25 (11) 0.38 � 0.22 (7) 0.82 Mann–Whitney U

WPL-inbredmale (multi

competition)*

0.29 � 0.15 (71) 0.50 � 0.23 (59) <0.001 Mann–Whitney U

Female

parasitization

lifespan (days)

WPL-inbred female* 14.13 � 3.79 (30) 10.92 � 4.39 (13) 0.05 Mann–Whitney U

WPL-outbred female* 19.98 � 6.35 (47) 15.17 � 6.35 (36) <0.001 Mann–Whitney U

Female

parasitization

(total hosts

parasitized)

WPL-inbred female* 49.07 � 15.14 (30) 26 � 16 (13) <0.001 Two sample t-test

WPL-outbred female* 80.79 � 26.75 (47) 40.56 � 24.08 (36) <0.001 Mann–Whitney U

Female

parasitization

(% hosts

parasitized out

of total offered)

WPL-inbred female* 69 � 13 (30) 44 � 19 (13) <0.001 Mann–Whitney U

WPL-outbred female* 79 � 14 (47) 53 � 20 (36) <0.001 Mann–Whitney U

Head width

(mm)

WPL-inbredmale* 0.699 � 0.030 (56) 0.732 � 0.049 (23) <0.001 Mann–Whitney U

WPL-inbred female* 0.686 � 0.059 (50) 0.768 � 0.052 (36) <0.001 Mann–Whitney U

WPL-outbred female 0.804 � 0.0563 (78) 0.795 � 0.067 (80) 0.43 Welch’s t-test

Whiting polyploid line ofNasonia vitripennis 661

unmated virgins that produced only sons). Of these 18

successful trials, the female mated with the haploidmale in

11 trials and with the diploid male in seven trials, which is

not a significant difference (Table 1).

In the multiple-mate competition assays, for 15 out of

20 trials, more than five out of the 10 females mated and

produced female offspring. Of these 15 trials, in six trials

the haploid male mated with a majority of the females, in

six trials the diploid mated with a majority of the females,

and in three trials the haploid male and the diploid male

mated with the same number of females (tied). Regardless

of whether the diploid or haploid male acquired more

mates in a specific trial, mating competition success ran-

ged widely for each male (i.e., 0–100% for haploids, 0–100% for diploids) (Table S1). In one multiple-mate com-

petition trial, all females were found to have mated with a

diploid male. As a low proportion of WPL red-eyed males

are diploid, it was possible that the red-eyed male used in

this trial as presumed haploid competitor was actually

diploid. Flow cytometry was used post-hoc to assess the

ploidy of the red-eyed male. It was confirmed a haploid;

therefore, this trial was retained in the analysis.

A binomial general linearized model was used to test

whether females were more likely to mate to a haploid

male or a diploid male. For the single-mate competition,

the 18 females were not more likely to mate with either

type of male, and the same applies for the 114 females of

the multiple-mate competition (Table S2). Taken

together, although sample sizes are limited, these results

do not point towards any difference in male mate compe-

tition ability between ploidy levels.

Progeny size and sex ratio

Diploid males do not sire more or fewer offspring than

haploid males. This applies for red-eyed males mated to

females in a no-choice experiment and those in single-fe-

male and multiple-female mate competition assays. In the

no-choice experiment, a WPL haploid red-eyed male and

a WPL diploid red-eyed male mated to an AsymCx female

had the same total progeny size (male and female offspring

of the female mate) and offspring sex ratio (males/total)

(Table 1, Figure 3).

In the single-mate competitions, an HVRx female

mated with either a WPL red-eyed haploid male or a WPL

Whiting inbred 2nWhiting inbred 3nWhiting outbred 2nWhiting outbred 3n

WPL 1n WPL 2n

Males Females

Lifespan (days)

Sur

viva

l pro

porti

on

Starved Starved

Fed Fed

A B

C D

Sur

viva

l pro

porti

on

0 2 4 6 8 10 0 2 4 6 8 10

0 10 20 30 40 50 0 10 20 30 40 50

Whiting inbred 2nWhiting inbred 3nWhiting outbred 2nWhiting outbred 3n

*

*

WPL 1n WPL 2n

Figure 2 Survival curves (proportion of individuals still alive over time) of non-polyploid and polyploid (A) starvedmales, (B) starved

females, (C) fedmales, and (D) fed females in the inbred and outbredWhiting polyploid line (WPL).Males assayed were inbred red-eyed

haploids and red-eyed diploids, females assayed were diploids and triploids. An asterisk denotes a significantly longer lifespan than the

counterpart. Note the different scales on the horizontal axes.

662 Leung et al.

wildtype-eyed diploid male. In 11 out of 18 trials, the

female mated with the haploid male. In the remaining

seven trials, the female mated with the diploid wildtype-

eyed male. They did not significantly differ in progeny size

or offspring sex ratio (Table 1). In a multiple-partner

experiment 10 HVRx females could be mated by a WPL

red-eyed haploid male or WPL wildtype-eyed diploid

male. Including the females that were not used in mate

competition analyses because a majority of females in their

trial did notmate, a total of 130 females mated, 71 with the

haploid male, and 59 with the diploid male (Table S1).

Progeny size is not significantly different between haploids

and diploids, but offspring sex ratio was significantly more

male-biased for diploid males (Table 1). For the multiple-

partner results the higher male offspring sex ratio for

females that mated with a diploid male is unusual, as it is

typical forNasonia spp. females to produce highly female-

biased broods under favorable conditions (Whiting,

1967).

Female parasitization rate

In the parasitization rate assay, WPL females were offered

10 fresh hosts every 2 days until death. Triploid females

lived significantly shorter and parasitized significantly

fewer hosts over their lifetime than diploid females (Fig-

ure 4). For the WPL-inbred females, diploid females lived

on average a significant 3.21 days longer then triploids

(Table 1). For the WPL-outbred females, diploid females

lived on average a significant 4.81 days longer (Table 1).

Survival distributions were also significantly different

between triploids and diploids for both backgrounds, with

a greater fraction of diploid females surviving at every time

point (log-rank test: v2 = 12.249, d.f. = 1, P<0.001) (Fig-ure 4A). Thus, when given a continuous supply of hosts,

used by the female both for oviposition and as a food

source, triploid females live significantly shorter than

diploid females. This contrasts somewhat with the results

of the ‘fed lifespan’ assay in which females was given 10%

sucrose solution, as there was no difference betweenWPL-

inbred diploid and triploid females, but WPL-outbred tri-

ploid females lived longer than diploid females (Fig-

ure 2D).

The diploid WPL-inbred females parasitized almost

twice as many hosts as their triploid counterparts, with a

parasitization rate (host parasitized/hosts offered) that was

a significant 25% higher (Table 1). The diploid WPL-out-

bred females also parasitized about double the number of

hosts relative to the outbred triploids, with a 26% higher

parasitization rate (Figures 4B and 4D). Interestingly,

although the inbred and outbred backgrounds cannot be

directly compared as independent experiments, the out-

bred background had much higher parasitization success

overall compared to the inbred background (for both the

diploid and triploid females) (Table 1).

A binomial general linearized mixed model was used to

analyze the relationship between ploidy and parasitization

success (Table S3). For all backgrounds, parasitization

ability declines over time as the female ages (Figure 4D),

but for the Whiting inbred background, controlling for

day (host set) and the individual, the diploid is 4.39more

likely to parasitize any given host than the triploid; the

diploid WPL-outbred female is 79 more likely to para-

sitize a host than the triploid WPL-outbred female

(Table S3).

Body size

Head width, used as proxy measure for total body

size, differed significantly between non-polyploid and

Pro

geny

size

(no.

Indi

vidu

als)

Offs

prin

g se

xra

tio (p

ropo

rtion

mal

es)

1n 2n 1n 2n 1n 2n

AsymCx HVRx single HVRx multiple mate

1.0

70

60

50

40

30

20

10

0

0.8

0.6

0.4

0.2

0.0

Female mate background

1n 2n 1n 2n 1n 2n

nsns

ns

ns

ns

*

A

B

Figure 3 Mean (� SD) (A) progeny size (no. familymembers)

and (B) offspring sex ratio (proportion sons) of haploid (1n) and

diploid (2n)Whiting inbred polyploid linemales with AsymCx

andHVRx female mates in the single- andmultiple-mate

competitions. The asterisk denotes a significant difference

between haploid vs. diploid fathers (Mann–Whitney U test:

P<0.0001; ns, P>0.05).

Whiting polyploid line ofNasonia vitripennis 663

polyploid individuals for two out of three cases

assessed. For WPL males, diploids are larger than

haploids (Table 1). For the WPL-inbred, diploid

females are smaller than triploids (Table 1). In con-

trast, diploid and triploid females of the WPL-outbred

are of similar size (Table 1).

Discussion

The consequences of inbreeding and polyploidy are well

known in CSD-species because of the sterile diploid male

vortex, but polyploid phenotypes in non-CSD species have

not been studied. In assaying a range of life-history traits

Figure 4 Female parasitization rate of

diploids (2n) and triploids (3n) of the

Whiting inbred and theWhiting outbred

backgrounds: (A) survival curve

(proportion of individuals still alive over

time), andmean (� SD) (B) total number

of hosts parasitized, (C) percentage of total

hosts offered parasitized, and (D) number

of hosts parasitized for each set of 10 fresh

hosts offered every 2 days until death. An

asterisk denotes a significant difference (in

panel A it denotes the significantly longer-

lived counterpart) (Mann–Whitney U test:

P<0.05).

664 Leung et al.

in a polyploid line of the non-CSD parasitoid N. vitripen-

nis, there appear to be a few disadvantages related to the

polyploid state (reproductive impairment, reduced

parasitization). These results can be interpreted for their

effects on the performance ofN. vitripennis as a biocontrol

agent, as it is used for muscid pest control in livestock rear-

ing (Kaufman et al., 2001a,b; Skovg�ard & Nachman,

2017), albeit uncommonly. But more importantly, the

results of this model species can be used to judge the

possible pros and cons of using polyploids in biocontrol

programs.

Lifespans of N. vitripennis WPL haploid and diploid

males (under both starved and fed conditions) are equal,

similar to results found for other hymenopteran species

such as CSD-species Habrobracon hebetor (Say), a para-

sitoid wasp (Clark & Rubin, 1961), and the ant Hypopon-

era opacior (Forel) (Kureck et al., 2013). This suggests that

hymenopteran diploid males are not impaired in longevity

compared to their haploid counterparts whether they have

CSD or not. For both ‘starved and fed lifespan’ assays, tri-

ploid females lived as long as diploid females of the WPL-

inbred line, but diploid females of the WPL-outbred line

lived longer than the triploids. The reason for this differ-

ence between backgrounds is unknown. It may be due to

co-adapted gene complexes being unbalanced in theWPL-

outbred cross relative to the WPL-inbred cross, which

would imply outbreeding depression. A shorter lifespan

may not necessarily be disadvantageous for individual fit-

ness or a biocontrol program if critical functions are con-

centrated in early life. For example, if fecundity peaks early

and reproductive output and parasitization ability is negli-

gible by the end of life, then a shorter lifespan may not

carry large fitness disadvantages. Notably, theWPL-inbred

background may confer a greater degree of starvation

resistance than the WPL-outbred, as both WPL-inbred

diploid and triploid females lived longer than their

WPL-outbred counterparts, a trend not observed under

fed conditions.

In both single- and multiple-mate competitions for

females of the HVRx background—a genetically variable

and recently derived laboratory population (van de Zande

et al., 2014) chosen specifically to circumvent multiple

mating, as wild outbred Nasonia spp. are strongly monan-

drous (van den Assem et al., 1980; Burton-Chellew et al.,

2007; Grillenberger et al., 2008)—, WPL haploids and

diploid males were equally successful in acquiring mates.

However, the small sample sizes and low statistical power

of this dataset should be noted. Owing to the labor inten-

siveness of this assay and the need to discard trials in which

females do not mate, a larger sample size was not possible.

However, we do not think that additional trials would

change results. In the Nasonia system, males present

courtship behaviors to females, which then have con-

trol over mate acceptance. Although the males of this

study were not evaluated for any specific differences in

courtship, the results imply that they may produce

similar olfactory, visual, and behavioral cues that are

not discriminated against by the females. Alternatively,

differences may exist and went unobserved, but did not

ultimately change the attractiveness of the diploid male.

This would be similar to, for example, the CSD-species

Cotesia glomerata (L.), in which diploid males initiate

courtship sooner than haploids, but have neither higher

nor lower mating success (Elias et al., 2009). The suc-

cess of diploid males in this study matches results for

haploid and diploid male mate competitions of other

CSD-parasitoid wasp species (reviewed by Harpur

et al., 2013; Thiel et al., 2014). This implies that the

utility of non-CSD diploid males in breeding programs

would not be hampered by unattractiveness, as even in

the presence of conventional haploid males, females are

receptive to mating with diploids.

When WPL haploid and diploid males were given

females of identical genetic background (AsymCx) in a

scenario in which there were no competing males, their

family size and offspring sex ratio were the same. The off-

spring sex ratio of those females that mated with the WPL

haploid and diploid males in no-choice assays were

approximately 0.15, which is not highly divergent from the

0.10 proportion that is possible when mated females have

ideal conditions (i.e., low competition, high host quality)

(Whiting, 1967). The diploid males of the WPL line are

thus exceptionally fertile relative to the near-universal

infertility of CSD diploid males (van Wilgenburg et al.,

2006; Heimpel & de Boer, 2008; Elias et al., 2009; Harpur

et al., 2013), although exceptions exist for the solitary ves-

pid wasp Euodynerus foraminatus (de Saussure) (Cowan &

Stahlhut, 2004), the ichneumonid wasp Diadromus pul-

chellusWesmael (El Agoze et al., 1994), and braconid para-

sitoids Cotesia vestalis (Haliday) (de Boer et al., 2007) and

C. glomerata (Elias et al., 2009). These CSD species with

fertile diploid males are capable of producing daughters.

The first two have haploid sperm and produce diploid

females, and the Cotesia spp. have diploid sperm and pro-

duce triploid females. In at least one other non-CSD spe-

cies (the braconid wasp Asobara japonica Belokobylskij)

diploid males are also capable of producing many triploid

daughters, although they do not sire as many offspring as

haploids (Ma et al., 2015). This may suggest that male

diploids in the non-CSD class have higher fertility than

CSD-species in general. Interestingly, whereas progeny size

was also similar between the WPL haploid and diploid

male in the mate competitions, offspring sex ratio of

diploid males was higher in the mate competition using

Whiting polyploid line ofNasonia vitripennis 665

genetically variable HVRx females (0.503, compared to

0.293 of haploid males).

Possible explanations for higher sex ratios among pro-

genies sired by diploid males are lower sperm count in

diploid males or impaired fertilization capacity of diploid

sperm. It was previously shown thatN. vitripennis haploid

males are able to mate with 13–18 females in rapid succes-

sion before being depleted of sperm (Beukeboom, 1994;

Chirault et al., 2016). These studies were the basis for

deciding to provide the males with 10 potential female

mates. It is possible that diploid males have fewer sperm

and experienced faster sperm depletion with successive

matings in the multiple-mate trials. Corroborating this, in

a study of haploid males, the most drastic reduction in

sperm transfer followed the first mating (Chirault et al.,

2016), which may explain why offspring sex ratio differ-

ences are not observed in the no-choice matings and sin-

gle-mate competitions. Diploid sperm may also be lower

in quality; for example, diploid sperm of CSD Habrobra-

con wasps apparently fail to penetrate eggs (MacBride,

1946). The results of this study are similar to those of other

parasitoid wasps (CSD braconids) where females did not

discriminate against diploid males as mates, but produced

fewer daughters (Elias et al., 2009; Thiel et al., 2014).

Regardless of why the male proportion is so high in proge-

nies of diploid males in the multiple-mate competition

experiments, this result suggests that diploid males of both

CSD and non-CSD species will have lower lifetime fitness

even if, as is the case with theWPL line, they live as long as

haploids and have high reproductive success with first

mates. This may severely reduce their usefulness in breed-

ing programs, but the degree of difference between diploid

and haploid lifetime fecundity has not been rigorously

tested. Follow-up investigation is needed to compare

N. vitripennis haploid and diploid male for lifetime fitness,

and to examine diploid sperm itself for impairment.

To the authors’ knowledge, the parasitization ability of

triploid female parasitoid wasps has not been tested for

any species previously, possibly because of their rarity.

Therefore, the WPL gives first insight on the parasitization

ability of polyploid parasitoid females. FemaleNasonia are

synovigenic, meaning that they emerge from their host

with a partial complement of mature eggs and can repro-

duce right away, but will continue to produce eggs as long

as they consume enough protein (Pannebakker et al.,

2013). Perhaps reflecting the better overall nutritional

composition of host hemolymph vs. the sucrose solution

of the ‘fed lifespan’ assays, females generally lived longer in

this assay (the exception being the WPL-diploids). This

matches observations of another chalcidoid wasp that is

possibly non-CSD, Aphytis melinus DeBach, living longer

and being more fecund with additional host feeding vs.

sugar feeding alone (Heimpel et al., 1997). However, the

parasitization assay of this study shows that female tri-

ploids both have shorter lives and are inferior lifetime par-

asitizers relative to diploids even with unlimited resources.

The hypothesis that triploids could be as proficient at kill-

ing hosts as diploids despite their low fecundity was based

on how Nasonia venom efficiently induces mortality in

hosts, even in the absence of any parasitoid feeding (Rivers

et al., 1993). As there was no evidence to suggest that the

venom of triploid females is attenuated, we predicted that

host killing ability might be retained. However, triploids of

both the WPL-inbred and WPL-outbred had poor para-

sitization performance in comparison to their diploid

counterparts. They parasitized far fewer hosts overall and a

lower percentage of offered hosts per set.

The underlying cause of reduced parasitization ability in

triploid females is not clear, but may possibly be attributed

to reduced fecundity. A large number of the triploid’s off-

spring die in the egg stage as aneuploids, so it is possible

that more hosts survived because of fewer offspring surviv-

ing to larval stage to consume hosts. Somewhat supporting

this, while parasitization gradually declined over the lifes-

pan of both diploids and triploids, parasitization would

increase and decrease with alternate host sets. This sinu-

soidal parasitization success pattern matches the pattern

expected for periodic egg depletion and replenishment as

is typical of female Nasonia synovigenic life-history strat-

egy, suggesting that every other host set may have had

more living offspring to kill hosts. Regrettably, as host kill-

ing was the focus of this assay, offspring were not counted.

Alternatively, it may be that triploids simply did not sting

as many hosts, and therefore failed to transfer venom as

often, or that they did not do as much host feeding as

diploids, which is also a factor in host killing (Kidd & Jer-

vis, 1989). Future studies should investigate these behav-

iors in triploids as possible explanatory factors for

parasitization rate deficiency, but these results suggest that

the triploid state is a major impediment to biological con-

trol efficiency.

Notably, an outbreeding advantage for parasitization

may have been observed in the females of this study. The

WPL-inbred cross has been used to maintain the line for

decades, so the cross with AsymCx females could be con-

sidered a single generation of outcrossing. Although the

assays for inbred and outbred females were run as separate

experiments, compared to theirWPL-inbred counterparts,

outbred diploid and triploid females parasitized more

hosts over their lifetime and parasitized a higher percent-

age of hosts. Impressively, the triploid outbred females

parasitized more than double the total number of hosts

parasitized by the inbred triploids (although they still par-

asitized only half as many hosts as the outbred diploids).

666 Leung et al.

Outbred lifespans were also approximately a third longer.

Lifespan extension with outbreeding is consistent with the

observation of Luna & Hawkins (2004), who found that

outbreeding can improve fecundity and lifespan in N. vit-

ripennis. These differences are too large to be accounted

for by any slight variation in host batch quality. Higher

parasitization rate overall for the WPL-inbred line could

have been from a larger number of offspring surviving to

larval stage to feed on the host, in which case the poor par-

asitization ability of inbred triploids was partly ‘rescued’

by outbreeding. Unfortunately, offspring data were not

collected and compared between the outbred and inbred

females because low and similar fecundity from aneu-

ploidy was expected for both, and should be considered in

future studies of triploid female performance.

Body size is an important factor in the overall fitness of

an insect. In general, larger insects outcompete smaller

conspecifics in reproductive output and resource competi-

tion (Beukeboom, 2018). In this study, the polyploid is

either not significantly different in body size from the non-

polyploid, or is only slightly larger. This is consistent with

studies on other invertebrates (reviewed by Fankhauser,

1945) and fits the overall trend for Hymenoptera (A Thiel,

pers. comm.). The diploid N. vitripennisWPL males were

slightly larger than the haploids according to the head

width proxy measurement, albeit with a large overlap in

values. This size difference did not affect any of the life-his-

tory traits assayed in this study. Similarly, in females, in

one background (WPL-outbred) the triploid females were

similarly sized as the diploids, and in the other background

(WPL-inbred), triploids were significantly larger. The rea-

son for this incongruency between backgrounds is not

known. It is possible that polyploidy in itself does not

cause major changes in body size for females, but is a con-

sequence of the scarlet mutant background used to main-

tain the WPL. Related to this, the WPL-outbred cross

could have counteracted an inbreeding effect on size

within the scarlet background. Both diploids and triploids

of this background are bigger than those of the inbred

background. This suggests that outbreeding can be used to

increase polyploid size, but the results here do not indicate

any practical advantage to doing so.

The results of this study can be interpreted in the con-

text of the use of non-CSD parasitoids for biological con-

trol. As polyploidy may be more viable in non-CSD

species, this suggests that advantageous polyploid applica-

tions in non-CSD parasitoid wasp breeding may be more

feasible than for CSD-species. It can, for example, hypo-

thetically assist with sex-specific tradeoffs, where traits

beneficial to females but detrimental to males would be

purged in the haploid state but be masked in the diploid

state, meaning that it can be passed on to the next

generation of females (the more important sex to biologi-

cal control). Our results do not support a promising role

of polyploidy in hymenopteran breeding, although there

may be some remedies to some of these drawbacks and

ways to improve the use of polyploids.

For possible diploid male fecundity loss from fewer or

less functional sperm, the number of diploid males can be

maximized to increase the net number of polyploid

females in the next generation. In Nasonia the develop-

ment of normal diploid embryos can be switched from

female tomale by silencing feminizing gene elements (Ver-

hulst, 2010). This can be exploited to produce families that

consist predominantly of diploid males. For females, the

problem of poor parasitization is harder to resolve, but the

proposed outbreeding solution can be tested in the future

to explore to what extent their biocontrol performance

can be optimized. In addition, artificial selection may be

applied to increase diploid male reproductive output and

triploid female parasitization ability. This study, however,

indicates a need for broader study of polyploid phenotypes

of non-CSD species, the problems that they can manifest,

and the specific solutions needed.

Acknowledgments

This study was funded by the Marie Curie Innovative

Training Network BINGO (Breeding Invertebrates for

Next Generation BioControl, project 641456). We thank

Yuan Zou, Elena Dalla Benetta, Marcela Bu�ri�cov�a, and

Anna Rensink for support in Nasonia culturing and data

collection, and are grateful to Peter Hoitinga for help with

statistical analyses and Andra Thiel for advising on

experimental design.

References

Asplen MK, Whitfield JB, de Boer JG & Heimpel GE (2009)

Ancestral state reconstruction analysis of hymenopteran sex

determination mechanisms. Journal of Evolutionary Biology

22: 1762–1769.van den Assem J & Jachmann F (1999) Changes in male persever-

ance in courtship and female readiness to mate in a strain of

the parasitic wasp Nasonia vitripennis over a period of 20+

years. Netherlands Journal of Zoology 49: 125–137.van den Assem J, Gijswijt MJ & N€ubel BK (1980) Observations

on courtship – and mating strategies in a few species of para-

sitic wasps (Chalcidoidea). Netherlands Journal of Zoology 30:

208–227.Beukeboom LW (1994) Phenotypic fitness effects of the selfish B

chromosome, paternal sex ratio (PSR) in the parasitic wasp

Nasonia vitripennis. Evolutionary Ecology 8: 1–24.Beukeboom LW (2018) Size matters in insects – an introduction.

Entomologia Experimentalis et Applicata 166: 2–3.

Whiting polyploid line ofNasonia vitripennis 667

Beukeboom LW & Desplan C (2003) Nasonia. Current Biology

13: R860.

Beukeboom LW&Kamping A (2006) No patrigenes required for

femaleness in the haplodiploid wasp Nasonia vitripennis.

Genetics 172: 981–989.Beukeboom LW & van de Zande L (2010) Genetics of sex deter-

mination in the haplodiploid waspNasonia vitripennis (Hyme-

noptera: Chalcidoidea). Journal of Genetics 89: 333–339.Beukeboom LW, Kamping A, Louter M, Pijnacker LP, Katju V

et al. (2007) Haploid females in the parasitic wasp Nasonia vit-

ripennis. Science 315: 206.

Breeuwer JAJ & Werren JH (1990) Microorganisms associated

with chromosome destruction and reproductive isolation

between two insect species. Nature 346: 558–560.de Boer JG, Ode PJ, Vet LEM,Whitfield JB & Heimpel GE (2007)

Diploid males sire triploid daughters and sons in the parasitoid

waspCotesia vestalis. Heredity 99: 288–294.Burton-Chellew MN, Beukeboom LW, West SA & Shuker

DM (2007) Laboratory evolution of polyandry in the para-

sitoid wasp Nasonia vitripennis. Animal Behaviour 74:

1147–1154.Charnov EL & Skinner SW (1984) Evolution of host selection

and clutch size in parasitoid wasps. Florida Entomologist 67:

5–21.Chirault M, van de Zande L, Hidalgo K, Chevrier C, Bressac C &

L�ecureuil C (2016) The spatio-temporal partitioning of sperm

by males of the prospermatogenic parasitoid Nasonia vitripen-

nis is in line with its gregarious lifestyle. Journal of Insect Physi-

ology 91–92: 10–17.Clark AM&RubinMA (1961) Themodification by X-irradiation

of the life span of haploids and diploids of the wasp,Habrobra-

con SP. Radiation Research 15: 244–253.Comai L (2005) The advantages and disadvantages of being poly-

ploid. Nature Reviews Genetics 6: 836–846.Cook JM (1993) Sex determination in the Hymenoptera: a review

ofmodels and evidence. Heredity 71: 421–435.Cook JM & Crozier RH (1995) Sex determination and popula-

tion biology in the Hymenoptera. Trends in Ecology & Evolu-

tion 10: 281–286.Cowan DP & Stahlhut JK (2004) Functionally reproductive

diploid and haploid males in an inbreeding hymenopteran

with complementary sex determination. Proceedings of the

National Academy of Sciences of the USA 101: 10374–10379.Dobson SL & Tanouye MA (1998) Evidence for a genomic

imprinting sex determination mechanism in Nasonia vitripen-

nis (Hymenoptera; Chalcidoidea). Genetics 149: 233–242.El Agoze M, Drezen JM, Renault S & Periquet G (1994) Analysis

of the reproductive potential of diploid males in the wasp

Diadromus pulchellus (Hymenoptera: Ichneumonidae). Bul-

letin of Entomological Research 84: 213–218.Elias J, Mazzi D & Dorn S (2009) No need to discriminate?

Reproductive diploid males in a parasitoid with complemen-

tary sex determination. PLoSONE 4: 1–7.Fankhauser G (1945) The effects of changes in chromosome

number on amphibian development. Quarterly Review of Biol-

ogy 20: 20–78.

Grillenberger BK, Koevoets T, Burton-Chellew MN, Sykes EM,

Shuker DM et al. (2008) Genetic structure of natural Nasonia

vitripennis populations: validating assumptions of sex-ratio

theory.Molecular Ecology 17: 2854–2864.Groothuis J & Smid HM (2017) Nasonia parasitic wasps escape

from Haller’s Rule by diphasic, partially isometric brain-body

size scaling and selective neuropil adaptations. Brain, Behavior

and Evolution 90: 243–254.Harpur BA, Sobhani M & Zayed A (2013) A review of the conse-

quences of complementary sex determination and diploidmale

production on mating failures in the Hymenoptera. Ento-

mologia Experimentalis et Applicata 146: 156–164.Heimpel GE & de Boer JG (2008) Sex determination in the

Hymenoptera. Annual Review of Entomology 53: 209–230.Heimpel GE, Rosenheim JA & Kattari D (1997) Adult feeding

and lifetime reproductive success in the parasitoid Aphytis

melinus. Entomologia Experimentalis et Applicata 83: 305–315.

Hein S, Poethke HJ & Dorn S (2009) What stops the ‘diploid

male vortex’? A simulation study for species with single locus

complementary sex determination. Ecological Modelling 220:

1663–1669.IBM (2017) IBM SPSS Statistics for Windows, v.25.0. IBM,

Armonk, NY, USA.

Kaufman PE, Long SJ & Rutz DA (2001a) Impact of exposure

length and pupal source onMuscidifurax raptorellus andNaso-

nia vitripennis (Hymenoptera: Pteromalidae) parasitism in a

New York poultry facility. Journal of Economic Entomology

94: 998–1003.Kaufman PE, Long SJ, Rutz DA & Waldron JK (2001b) Para-

sitism rates ofMuscidifurax raptorellus and Nasonia vitripennis

(Hymenoptera: Pteromalidae) after individual and paired

releases in New York poultry facilities. Journal of Economic

Entomology 94: 593–598.Kidd N & Jervis M (1989) The effects of host-feeding behavior on

parasitoid-host interactions, and the implications for biologi-

cal control. Researches on Population Ecology 31: 235–274.Kureck IM, Nicolai B & Foitzik S (2013) Similar performance of

diploid and haploid males in an ant species without inbreeding

avoidance. Ethology 119: 360–367.van Lenteren JC (2012) The state of commercial augmentative

biological control: plenty of natural enemies, but a frustrating

lack of uptake. BioControl 57: 1–20.van Lenteren JC, Roskam MM & Timmer R (1997) Commercial

mass production and pricing of organisms for biological con-

trol of pests in Europe. Biological Control 10: 143–149.van Wilgenburg E, Driessen G & Beukeboom LW (2006) Single

locus complementary sex determination in Hymenoptera: an

‘unintelligent’ design? Frontiers in Zoology 3: 1–15.LunaMG&Hawkins BA (2004) Effects of inbreeding versus out-

breeding inNasonia vitripennis (Hymenoptera: Pteromalidae).

Environmental Entomology 33: 765–775.Ma WJ, Pannebakker BA, van de Zande L, Schwander T,

Wertheim B & Beukeboom LW (2015) Diploid males support

a two-step mechanism of endosymbiont-induced thelytoky in

a parasitoid wasp. BMCEvolutionary Biology 15: 84.

668 Leung et al.

MacBride DH (1946) Failure of sperm of Habrobracon diploid

males to penetrate the eggs. Genetics 31: 224.

Otto SP & Whitton J (2000) Polyploid incidence and evolution.

Annual Review of Genetics 34: 401–437.Pannebakker BA, Trivedi U, Blaxter MA, Watt R & Shuker DM

(2013) The transcriptomic basis of oviposition behaviour in

the parasitoid waspNasonia vitripennis. PLoS ONE 8: e68608.

Piferrer F, Beaumont A, Falgui�ere JC, Flaj�shans M, Haffray P &

Colombo L (2009) Polyploid fish and shellfish: production,

biology and applications to aquaculture for performance

improvement and genetic containment. Aquaculture 293: 125–156.

R Core Team (2014) R: A Language and Environment for Statisti-

cal Computing. R Foundation for Statistical Computing,

Vienna, Austria.

Rivers DB, Hink WF & Denlinger DL (1993) Toxicity of the

venom fromNasonia vitripennis (Hymenoptera: Pteromalidae)

toward fly hosts, nontarget insects, different developmental

stages, and cultured insect cells. Toxicon 31: 755–765.Shuker D, Lynch J & Morais AP (2003) Moving from model to

non-model organisms? Lessons fromNasoniawasps. BioEssays

25: 1247–1248.Shuker DM, Phillimore AJ, Burton-Chellew MN, Hodge SE

& West SA (2007) The quantitative genetic basis of poly-

andry in the parasitoid wasp, Nasonia vitripennis. Heredity

98: 69–73.Skovg�ard H & Nachman G (2017) Modelling Biological Control

of Stable Flies by Means of Parasitoids. Danish Environmental

Protection Agency, Copenhagen, Denmark.

Stouthamer R, Luck RF & Werren JH (1992) Genetics of

sex determination and the improvement of biological con-

trol using parasitiods. Environmental Entomology 21: 427–435.

Thiel A,Weeda AC& Bussi�ere L (2014) Haploid, diploid, and tri-

ploid – discrimination ability against polyploid mating partner

in the parasitic wasp, Bracon brevicornis (Hymenoptera: Bra-

conidae). Journal of Insect Science 14: 1–7.Verhulst EC (2010) Maternal control of haplodiploid sex deter-

mination in the waspNasonia. Science 328: 620–623.Verhulst EC, Lynch JA, Bopp D, Beukeboom LW& van de Zande

L (2013) A new component of the Nasonia sex determining

cascade ismaternally silenced and regulates transformer expres-

sion. PLoS ONE 8: e63618.

Werren JH, Richards S, Desjardins CA, Niehuis O, Gadau J et al.

(2010) Functional and evolutionary insights from the genomes

of three parasitoidNasonia species. Science 327: 343–348.

Wertheim B, Beukeboom LW & van de Zande L (2013) Poly-

ploidy in animals: effects of gene expression on sex determina-

tion, evolution and ecology. Cytogenetic and Genome

Research 140: 256–269.Whiting PW (1960) Polyploidy inMormoniella. Genetics 7: 949–970.

Whiting AR (1967) The biology of the parasitic wasp Mor-

moniella vitripennis [= Nasonia brevicornis] (Walker). Quar-

terly Review of Biology 42: 333–406.van de Zande L, Ferber S, de Haan A, Beukeboom LW, van Heer-

waarden J & Pannebakker BA (2014) Development of a Naso-

nia vitripennis outbred laboratory population for genetic

analysis. Molecular Ecology Resources 14: 578–587.Zaviezo T, Retamal R, Urvois T, Fauvergue X, Blin A & Malausa

T (2018) Effects of inbreeding on a gregarious parasitoid wasp

with complementary sex determination. Evolutionary Applica-

tions 11: 243–253.Zayed A & Packer L (2005) Complementary sex determination

substantially increases extinction proneness of haplodiploid

populations. Proceedings of the National Academy of Sciences

of the USA 102: 10742–10746.

Supporting Information

Additional Supporting Information may be found in the

online version of this article:

Table S1.Winners of single- and multiple-mate compe-

titions of Whiting inbred lines for HVRx females. Grayed-

out trials were discarded from analyses because the female

of single competition did not mate, or a majority of

females in amultiple-mate competition did not

Table S2. Results from the general linearized model for

likelihood of females to mate with a haploid or diploid

male in the multiple-mate experiment. The intercept indi-

cates the model incorporating trial number as a random

factor. The reference category is set to the haploidmale

Table S3. Results from the binomial general linearized

mixed model for parasitization rate of inbred and outbred

Whiting line females. The intercept indicates the model

incorporating random factors ‘day’ (host set) and ‘individ-

ual’, and the fixed effect ‘ploidy’ or ‘background’. The data

for the diploid are relative to the triploid (0a) reference cat-

egory for failed parasitization

Whiting polyploid line ofNasonia vitripennis 669